A Phase II Evaluation of the Efficacy and Safety of Sacituzumab Govitecan in Patients with Recurrent Uterine Cancer

Abstract

Purpose:

Patients with endometrial cancer who progress after chemotherapy/immunotherapy have limited treatment options. We evaluated the activity and safety of sacituzumab govitecan (SG), a Trop-2–directed antibody–drug conjugate, in patients with advanced/recurrent endometrial cancer, including carcinosarcoma.

Patients and Methods:

This was a phase II, two-stage, open-label, investigator-initiated trial of patients with persistent/recurrent endometrial cancer who had progressed following ≥1 prior chemotherapy. Patients received SG 10 mg/kg on days 1 and 8 every 3 weeks. The primary endpoint was objective response rate (ORR) by RECIST v1.1. Secondary endpoints included clinical benefit rate [CBR = complete response (CR) + partial response (PR) + stable disease ≥ 6 months], duration of response (DOR), progression-free survival (PFS), overall survival (OS), and safety. Trop-2 expression was analyzed by immunohistochemistry as an H-score.

Results:

Fifty patients were screened, and 21 enrolled during stage 1; 34 patients were screened, and 29 enrolled during stage 2; 84% (n = 42) of the patients harbored serous carcinoma, carcinosarcoma, or grade 3 endometrioid tumors. Patients received a median of two prior therapies (range, 1–4) and 50% had failed pembrolizumab/dostarlimab. At a median follow-up (range) of 11 (2.9–65.5) months, the ORR was 28% [95% confidence interval (CI), 16%–42%], including CRs (4%) and PRs (24%). The median DOR (95% CI) was 9.3 (2–12.9) months, with four still responding. The CBR was 52% (26/50). The median PFS and OS were 5.5 (95% CI, 3.7–7.4) and 17.5 (95% CI, 10.4–22.2) months, respectively. Grade 3 to 4 toxicity occurred in 88% with no attributable deaths. Mean H-scores did not predict response.

Conclusions:

SG demonstrated encouraging efficacy in a pretreated population that included biologically aggressive recurrent endometrial cancer. Adverse events were consistent with the known safety profile.

Translational Relevance.

Patients with advanced/recurrent endometrial cancer who progress after platinum-based and checkpoint inhibitor therapy have limited options. NCT04251416 evaluated the efficacy/safety of sacituzumab govitecan (SG), a Trop-2–directed antibody–drug conjugate, in advanced/recurrent endometrial cancer. Mean membrane H-score did not vary with objective response, but there was moderate correlation between maximum change in tumor volume and H-score. Objective responses were achieved in 28% (14/50) and were not limited to a particular histology. SG had a manageable safety profile; adverse events were consistent with the known safety profile of SG.

Introduction

Annually, there are nearly half a million new cases of endometrial cancer worldwide (1). Endometrial cancer is the most common gynecologic malignancy in the United States, with 69,120 new cancer cases and 13,860 deaths expected in 2025 in the United States alone (2). Endometrial cancer incidence is projected to double by 2030 (3).

For recurrent disease or advanced cases with measurable disease, the addition of anti-PD1 immunotherapy to carboplatin and paclitaxel has recently been shown to offer clear clinical benefit (4, 5). Therapeutic options for patients with recurrent disease who fail platinum and checkpoint inhibition remain limited (6). Uterine cancer also suffers from a low funding-to-lethality ratio which may hinder the development of effective agents for the disease (7). Among a database of 1,961 patients with advanced or recurrent endometrial cancer diagnosed in the United States between 2013 and 2021 who received at least one line of therapy largely in community cancer centers, the median overall survival (OS) ranged from only 14.1 to 31.8 months. The median duration of second-line therapy was as short as 3.2 months; only 27% of patients received a third line of treatment, likely reflecting in part the paucity of meaningful approaches for this disease (8).

Trop-2 (trophoblast cell surface antigen 2 or TACSTD2) is a transmembrane glycoprotein overexpressed in a variety of gynecologic malignancies, including serous carcinomas (9) and carcinosarcomas (10). Its expression regulates numerous growth factor pathways and portends poor prognosis (11). Trop-2 has thus emerged as a relevant therapeutic target. Sacituzumab govitecan (SG) is a Trop-2–directed antibody–drug conjugate (ADC) composed of a Trop-2–targeted antibody conjugated with SN-38, an active metabolite of irinotecan (a topoisomerase I inhibitor), via a proprietary hydrolyzable linker (12, 13) that achieves a drug to antibody ratio (DAR) of 8:1 (14). This is higher than many other ADCs used in gynecologic oncology (e.g., tisotumab vedotin: 4, mirvetuximab soravtansine: 3.5, trastuzumab deruxtecan: 8; refs. 15–17). Higher DARs offer the potential for greater therapeutic potency; however, in a multivariate analysis of ADCs from 2014 to 2024, DARs >4 also portend hematologic toxicity (18).

SG has shown antitumor activity in multiple solid tumors and encouraging preclinical activity against poorly differentiated endometrial cancer cell lines overexpressing Trop-2 (19–22). SG has been approved for unresectable, locally advanced or metastatic breast cancer (23). In the phase 1/2 IMMU-132-01 basket trial, SG demonstrated a 22% objective response rate (ORR) and OS of 11.9 months in patients with endometrial cancer who had relapsed after or were refractory to at least one prior standard treatment regimen (24). In the recently published TROPiCS-03 (ClinTrials.gov ID: NCT03964727) trial, a multicohort, open-label, phase II basket study in patients with metastatic solid tumors, SG demonstrated an ORR of 22% in 41 patients with persistent/recurrent endometrial cancer (25).

We conducted a single center, open-label, phase II, two-stage study that evaluated the efficacy and safety of SG in patients with recurrent endometrial cancer with measurable disease (ClinTrials.gov ID: NCT04251416). The first stage of the study selected for patients with Trop-2 expression; the second stage was unenriched. Herein, we report on patients enrolled in both the first and second stages of NCT04251416, who had progression after prior platinum-based chemotherapy with or without anti-PD1–directed therapy. The present study complements the findings of TROPiCS-03 by providing information about the response of high-grade tumors such as serous carcinomas and carcinosarcomas to SG in a more ethnically diverse patient population.

Patients and Methods

Study design and patient population

This was a single center, open-label, phase II study, assessing the efficacy and safety of SG in patients with persistent/recurrent endometrial cancer. Patients were enrolled between March 2020 and December 2024. Patients had to have histologically documented advanced or metastatic endometrial cancer that had progressed after at least one prior platinum-based line of chemotherapy. Key inclusion criteria included an Eastern Cooperative Oncology Group performance status (RRID: SCR_026432) of 0 to 1, creatinine clearance ≥30 mL/minute, adequate marrow function, age ≥18 years, and no limit on prior lines of therapy. Key exclusion criteria included leiomyosarcoma and/or endometrial stromal sarcomas. Full eligibility criteria are included in the protocol (Supplementary Data S1). The first stage of enrollment enriched for patients with Trop-2 expression of at least 2+ staining intensity of Trop-2 in 50% of tumor cells (SP295 antibody, Spring Bioscience Co; RRID: AB_3075505); however, in line with FDA approvals of SG in breast cancer without need for a companion diagnostic (23) and the observation that >90% of endometrial cancers tested in stage 1 exhibited Trop-2 expression, recruitment was subsequently expanded in the second stage without the need for prospective Trop-2 testing. Exploratory analyses (i.e., H-score) of Trop-2 expression were then conducted using a newly developed antibody (EPR20043, Abcam; RRID: AB_2811182) with similar specificity but superior staining intensity (internal data, Gilead) for both stages 1 and 2 tumors. The new clone was chosen by the diagnostic assay manufacturer (Roche Tissue Diagnostics; RRID: SCR_013652) as the preferred clone for future companion diagnostic development. Written informed consent was provided by all patients. The study protocol was approved by the local Institutional Review Board and conducted in accordance with the Declaration of Helsinki and the International Conference on Harmonization Good Clinical Practice guidelines.

Treatment

Patients received 10 mg/kg of SG via intravenous infusion on days 1 and 8 of a 21-day cycle until progressive disease (PD), unacceptable toxicity, study withdrawal, or death, whichever occurred first. Drug was supplied by Gilead Sciences, Inc.

Endpoints

The primary endpoint was ORR [confirmed complete response (CR) and partial response (PR)] by investigator’s assessment per RECIST v1.1 (RRID: SCR_026435; ref. 26). Secondary endpoints included clinical benefit rate (CBR), duration of response (DOR), progression-free survival (PFS) by investigator’s assessment per RECIST v1.1, OS, and safety. Exploratory endpoints included correlation of ORR, CBR, DOR, and PFS with Trop-2 expression levels.

CBR was defined as best response of CR + PR + stable disease (SD) for at least 6 months, according to RECIST v1.1 by investigator’s assessment. DOR was measured from the date of first response (CR or PR) to the date of first documentation of disease progression from any cause as assessed by the investigator per RECIST v1.1 or death, whichever occurred first; patients not known to have progressed or died were censored on the date of last tumor assessment. Only the patients with confirmed CR and PR were included in the analysis of DOR. PFS was measured from the date of the first dose of study drug to the date of the first documentation of disease progression by investigator’s assessment based on RECIST v1.1 or death due to any cause, whichever came first; patients not known to have progressed or died were censored on the date of last tumor assessment. OS was measured from the date of the first dose of study drug to death from any cause; patients not known to have died were censored on the date they were last known to be alive. The frequency and severity of adverse events (AE) were classified by Common Terminology for Adverse Events v 5.0 (RRID: SCR_010296).

Statistical analysis

To make informed decisions about treatment efficacy early in the trial, the protocol used a Simon optimal two-stage design but did not suspend accrual between stages. Briefly, 21 eligible and evaluable patients were enrolled into the first stage; if 3 or more of them responded (i.e., CR or PR), then accrual to stage 2 would continue to enroll 29 additional patients to yield 50 eligible and evaluable patients. Upon full accrual, eight or more responses would allow rejection of the null hypothesis ORR = 10%, and the regimen would be considered worthy for additional investigation. If the true ORR = 25%, then this design has power = 90.08% at one-sided α = 10% to detect it. DOR, PFS, and OS estimates were calculated using the Kaplan–Meier method. Confidence intervals (CI) for medians were constructed with the method of Brookmeyer and Crowley (27). Exact binomial CIs using the Clopper–Pearson method (28) were calculated for ORR and CBR. All CIs, including that for the primary endpoint of ORR, are two-sided 95% CIs in accordance with conventional reporting practices. Histologies were compared for ORR differences with Fisher's exact test. The Cox proportional hazards model was used in efficacy analyses by Trop-2 subgroup. All statistical analyses were performed using SAS (SAS Institute, v9.2 or later; RRID: SCR_008567) or R (v4.0.5; RRID: SCR_001905).

Analysis of Trop-2 expression

Stage 1 selected for tumors with Trop-2 expression with 2+ staining using SP295 antibody at Roche Tissue Diagnostics. Stage 2 was open to all-comers. Exploratory analyses of Trop-2 expression in both stage 1 and 2 tumors were subsequently performed on formalin-fixed, paraffin-embedded archival tumor specimens using the anti–Trop-2 EPR20043 clone due to its enhanced intensity relative to SP295 (Gilead internal data; ref. 29). Immunohistochemistry (IHC) staining was defined as 0 (no staining), 1+ (weak staining), 2+ (moderate staining), or 3+ (strong staining). Membrane H-scores on a scale of 0 to 300 were calculated using the following formula: (1 × % cells 1+) + (2 × % cells 2+) + (3 × % cells 3+; refs. 30, 31). H-score is frequently used in the development of IHC assays for biomarkers for exploratory analyses where the scoring method and cutoff have not been established. Trop-2 scoring was conducted using a single sample per participant by a pathologist using either primary or metastatic disease. Informed consent was obtained for the evaluation of Trop-2 expression. Membrane scores represent staining limited to the cellular membrane. Kruskal–Wallis tests were used in the comparison of IHC parameters with response.

Results

Study participants

At the data extraction date of September 19th, 2025, 50 patients were evaluable. A total of 50 patients were screened for stage 1, yielding 21 evaluable; a total of 34 patients were screened for stage 2, yielding 29 evaluable (Supplementary Data S2). All patients received ≥1 dose of SG; there were 42 progression-free events at this time, and 50 patients were evaluable for CBR. The median follow-up was 11 months (range, 2.9–65.5 months). Demographic and disease characteristics are shown in Table 1. The median age was 68 years (range, 30–82), and patients received a median of two (range, 1–4) prior anticancer regimens. All patients had received carboplatin/paclitaxel, 50% had received pembrolizumab (n = 21) or dostarlimab (n = 4), and 26% (n = 13) had received doxorubicin. Overall, 24 (48%) patients had serous (mixed or pure) carcinoma, 11 (22%) had carcinosarcoma, 11 (22%) had endometrioid histology, and 4 (8%) had other histologies (dedifferentiated, mesonephric, and gastrointestinal). Among the endometrioid tumors, most (73%%, 8/11) were grade 3, 18% (2/11) were grade 2, and 9% (1/11) were grade 1. Mismatch repair (MMR) status was available in 46 patients; only 4.3% (one endometrioid grade 2 and one dedifferentiated tumor) were MMR-deficient, both with loss of MLH1 and PMS2. A total of 76% (n = 38) patients had advanced-stage disease at initial diagnosis.

Table 1.

Patient and disease characteristics.

Characteristic	n	%
Age (years)	​	​
<65 years	32	64
≥65 years	18	36
Race	​	​
Non-Hispanic White	33	66
Non-Hispanic Black	7	14
Non-Hispanic Asian	2	4
Non-Hispanic other	3	6
Hispanic	5	10
Disease status	​	​
Recurrent	36	72
Refractory	14	28
ECOG performance status	​	​
0	27	54
1	23	46
FIGO (2008) stage	​	​
IA	6	12
IB	4	8
II	2	4
IIIA	5	10
IIIB	1	2
IIIIC1	4	8
IIIC2	7	14
IVA	4	8
IVB	17	34
Histology	​	​
Endometrioid	11	22
Grade 1	1	2
Grade 2	2	4
Grade 3	8	16
Mixed serous	3	6
Serous	21	42
Carcinosarcoma	11	22
Other	4	8
MMR status (n = 46)	​	​
Proficient	44	96
Deficient	2	4
Prior treatments	​	​
Doxorubicin	13	26
Pembrolizumab	21	42
Other	8	16

​	Median	Range
Prior lines of chemotherapy	2	1–4

​	Mean	SD
Membrane Trop-2 H-score (n = 43)	​	​
All	212.4	75.2
Nonresponders	198.4	80.5
Responders	244.8	49.9
​	P = 0.072

Abbreviations: ECOG, Eastern Cooperative Oncology Group; FIGO, Federation Internationale d’Obstetrique et Gynecologie; MMR, mismatch repair; SD, standard deviation.

Efficacy analysis

Confirmed objective response was observed in 14 patients (28%; 95% CI, 16%–42%), including 2 (4%) CRs and 12 (24%) PRs. SD was observed in 26 (52%) patients, and 10 (20%) had PD. CBR (95% CI) was 52% (37%–66%) based on 26 of 50 evaluable (Table 2). Objective responses were not limited to a particular histologic type (P = 0.32). Specifically, an objective response was achieved in 33% (7/21) of serous, 36% (4/11) of endometrioid, 9% (1/11) of carcinosarcoma, 0% (0/3) mixed serous tumors, and 50% (2/4) of ‘other’ histologies. Among these 14 responders, the median DOR (95% CI) was 9.3 (95% CI, 2–12.9) months as of 09/30/2025, with four still responding. The median PFS (95% CI) was 5.5 (3.7–7.4) months; the median OS (95% CI) was 17.5 (10.4–22.2) months (Fig. 1). A total of 70.8% (34/48) patients had a reduction in target lesion diameter [Fig. 2 (left)]; many of the responses were durable [Fig. 2 (right)]. Of note, two patients with PD are not represented due to clinical progression prior to planned imaging assessment. ORR for stages 1 and 2 were 33% and 25%, respectively.

Table 2.

Response characteristics.

​	All patients (n = 50)
ORR (confirmed CR + PR), n (%; 95% CI)	14 (28%; 16%–42%)
Best overall response, n (%)	​
CR	2 (4%)
PR	12 (24%)
SD	26 (52%)
PD	10 (20%)
Not evaluable/not assessed	0
Median DORa (95% CI), months	9.3 (2–12.9)
CBR (confirmed CR + PR + SD ≥ 6 months), n (%; 95% CI)	26 (52%; 37%–66%)

^aAssessed in the n = 14 participants who achieved CR or PR.

Figure 1.

Progression-free (PFS, top) and overall (OS, bottom) survival with 95% CIs.

Figure 2.

Waterfall plot (left) illustrating best percentage change from baseline in total sum of target lesion diameters and spider plot (right) of tumor response over time.

Trop-2 expression

Tumor IHC was evaluable in 43 patients (86%). There were seven participants without biospecimens for the calculation of H-score; in four instances there was insufficient viable tumor for analysis, and in three instances, the specimen could not be obtained. Using anti–Trop-2 antibody EPR20043, the mean ± standard deviation membrane H-score among all patients was 212.4 ± 75.2 and did not vary with objective response (198.4 ± 80.5 vs. 244.8 ± 49.9, P = 0.07; Fig. 3 (top)). There was a large negative correlation (P = −0.45) between maximum change in tumor volume and H-score [P = 0.003; Fig. 3 (middle)]. The median H-score was 240 (Fig. 3 (bottom)). The median PFS among patients with H-scores below the median was 5.6 months (95% CI, 3.5–9.2), versus 3.8 months (95% CI, 3.4–11.3) for patients at or above the median (HR 0.93; 95% CI, 0.47–1.81, P = 0.83). There was a 40% (9/22) ORR in patients with H-scores at or above the median, versus 19% (4/21) in patients with H-scores below the median (P = 0.19). As most patients (92%) demonstrated some level of tumor Trop-2 expression based on an H-score greater than 0 in stage 1, the protocol was amended before opening enrollment to the second stage to remove Trop-2 prescreening testing. Previous work (25) has shown that patients with CR, PR, SD, and PD exhibit similar ranges of H-scores among and no differences in PFS based on median H-score, suggesting a limited association between response to SG and Trop-2 expression.

Figure 3.

IHC characteristics and response (n = 43). Archival tumor was stained with the anti–Trop-2 EPR20043 antibody. Membrane H-score versus best response (top) or tumor volume change at best response (middle). Kaplan–Meier estimates of PFS by median H-score (bottom).

Safety analysis

Grade 3 to 4 treatment-emergent AEs (TEAE) were reported 323 times in 44 (88%) patients (Table 3). The most common grade 3 to 4 TEAEs were neutropenia (80 instances among n = 26 patients, 52%), anemia (38 instances among n = 19 patients, 38%), and fatigue (14 instances among n = 10 patients, 20%). Serious AEs occurred 85 times among 25 (50%) patients, most commonly neutropenia (8 instances among n = 7 patients, 14%), febrile neutropenia (5 instances among n = 5 patients, 10%), sepsis (5 instances among n = 4 patients, 8%), and diarrhea (4 instances among n = 4 patients, 8%; Supplementary Data S3). Toxicities were aligned with the known safety profile for the drug when used in other solid tumors. AEs were overcome by using readily available approaches such as growth factor injections and antimotility agents such as loperamide along with supportive care fluid and electrolytes, as described in the FDA-approved product monograph. Three patients discontinued treatment for grade 3 to 4 TEAEs, including myocardial infarction (n = 1), cerebral edema (n = 1), and pleural effusion/hypoxia/dyspnea (n = 1). Treatment delays and dose reductions were required in 27 and 9 patients, respectively. The small samples size and lack of a control arm prevented the ability to draw conclusions between AEs and Trop-2 expression. Given that grade 3 events occurred in 88% (n = 44/50) of patients, it is assumed that toxicity was overall consistent across Trop-2 levels.

Table 3.

TEAEs, grades 3–4.

System organ class (category)	CTCAE term	Number affected	% affected	TEAEs
Blood/lymphatic system	Anemia	19	38%	38
​	Other	6	12%	22
​	Febrile neutropenia	6	12%	6
Cardiac disorders	Myocardial infarction	1	2%	1
​	Ventricular tachycardia	1	2%	1
Gastrointestinal disorders	Abdominal pain	1	2%	1
​	Colitis	1	2%	1
​	Colonic obstruction	2	4%	2
​	Diarrhea	5	10%	5
​	Other	1	2%	1
​	Nausea	1	2%	1
​	Small intestinal obstruction	1	2%	1
General disorders/administration site conditions	Fatigue	10	20%	14
​	Pain	1	2%	1
Infections/infestations	Bronchial infection	1	2%	1
​	Enterocolitis infection	1	2%	1
​	Other	3	6%	3
​	Lung infection	2	4%	2
​	Sepsis	4	8%	5
​	Skin infection	1	2%	1
​	Urinary tract infection	1	2%	1
Injury/poisoning/procedural complications	Fall	2	2%	1
​	Fracture	3	6%	3
Investigations	Alkaline phosphatase increased	1	2%	1
​	Blood bilirubin increased	2	4%	2
​	Cardiac troponin increased	1	2%	1
​	Creatinine increased	1	2%	1
​	Lymphocyte count decreased	3	6%	32
​	Neutrophil count decreased	26	52%	80
​	Platelet count decreased	1	2%	1
​	White blood cell decreased	3	6%	34
Metabolism/nutrition disorders	Anorexia	1	2%	1
​	Dehydration	1	2%	1
​	Hypoalbuminemia	1	2%	1
​	Hypocalcemia	2	4%	2
​	Hypophosphatemia	4	8%	8
​	Other	1	2%	2
Musculoskeletal/connective tissue	Arthralgia	1	2%	1
​	Flank pain	1	2%	1
​	Generalized muscle weakness	3	6%	3
​	Joint effusion	1	2%	2
​	Myalgia	1	2%	1
​	Pain in extremity	2	4%	2
Neoplasms benign and malignant	Other	1	2%	1
Nervous system disorders	Cerebral edema	1	2%	1
​	Syncope	2	4%	2
Psychiatric disorders	Agitation	1	2%	1
Renal/urinary disorders	Acute kidney injury	4	8%	5
​	Renal calculi	1	2%	1
Respiratory/thoracic/mediastinal disorders	Cough	1	2%	1
​	Dyspnea	3	6%	3
​	Hypoxia	2	4%	3
​	Pleural effusion	1	2%	1
​	Other	2	4%	2
Skin/subcutaneous tissue disorders	Rash, maculopapular	1	2%	1
​	Other	2	4%	2
Vasular disorders	Hypertension	1	2%	2
​	Hypotension	3	6%	3
​	Thromboembolic event	2	4%	3

Abbreviation: CTCAE, Common Terminology for Adverse Events.

Discussion

In this phase II study, SG monotherapy showed an encouraging ORR of 28% in patients with advanced endometrial cancer who failed carboplatin/paclitaxel (100%), anti-PD1 (50%), and doxorubicin (26%). The study met its primary efficacy endpoint, ORR >10%. The ORR is particularly encouraging given the representation of biologically aggressive histologic subtypes (84%), including serous (pure and mixed) carcinoma, grade 3 endometrioid, or carcinosarcoma, the latter of which is commonly excluded from trial designs, with no limit on prior lines of therapy or treatment refractoriness. The ORR compares very favorably to historic rates for single-agent chemotherapies, including the most active agent, paclitaxel (27.3%; ref. 32). Doxorubicin, though active as a first-line agent, has exhibited a response rate as low as 0% in pretreated patients (33, 34); responses for docetaxel (35), ixabepilone (36), liposomal doxorubicin (37), oxaliplatin (38) are 7.7%, 12%, 9.5%, and 13.5%, respectively. The activity of SG as monotherapy is also respectable relative to modern combination approaches. For example, in the phase III KEYNOTE-775 trial, 827 patients with advanced endometrial cancer who had disease progression after platinum-based chemotherapy were randomized to receive lenvatinib plus pembrolizumab (n = 411) or single-agent chemotherapy (n = 416). With single-agent chemotherapy, the ORR was 15%, median PFS was 3.8 months, median OS was 11.4 months, and median DOR was 5.7 months; results for the lenvatinib–pembrolizumab arm were 32%, 7.2, 18.3, and 14.4 months, respectively. Notably, the patients in KEYNOTE-775 had not received an immune checkpoint inhibitor (ICI), and the majority were treated in the second line (39). The patient population in the present trial was older, more heavily pretreated, and included a larger percentage of difficult-to-treat histologies. Similarly, in a phase I/II study of 28 patients with platinum-refractory or -resistant endometrial cancer, novel combination chemotherapy with gemcitabine, levofolinate, irinotecan, and 5-fluorouracil showed an ORR of only 7%; the median PFS was 3 months, and the median OS was 12 months (40).

The side effect profile of SG in endometrial cancer mirrors that described in the phase III ASCENT trial (NCT02574455), which compared SG versus physician’s choice chemotherapy in metastatic triple-negative breast cancer. In breast cancer, despite higher rates of grade $\mathrm{\ge }$ 3 neutropenia (51%), diarrhea (10%), leukopenia (10%), anemia (8%), and febrile neutropenia (6%; ref. 41), health-related quality of life (HRQoL) as assessed using European Organisation for Research and Treatment of Cancer QLQ-C30 demonstrated a longer median time to first worsening for SG in several areas: pain (21.6 vs. 9.9 weeks, P < 0.001), physical functioning (22.1 vs. 12.1 weeks, P < 0.001), role functioning (11.4 vs. 7.1 weeks, P < 0.001), and fatigue (7.7 vs. 6 weeks, P < 0.05). Changes from baseline also favored SG over chemotherapy for global health status/QoL, physical functioning, fatigue, and pain; they were noninferior in all other domains except nausea/vomiting/diarrhea (42). Notably, in the present phase II study, AEs were common but anticipated and manageable with established supportive care measures. HRQoL was not captured but would represent an opportunity for additional future characterization.

Findings from the current study are consistent with the results from the phase II TROPiCS-03 endometrial cancer cohort basket study in 41 patients with recurrent endometrial cancer, in which SG monotherapy resulted in an ORR of 22%; the median DOR was 8.8 months (95% CI, 2.8 to not estimable), and the median PFS was 4.8 months (95% CI, 2.8–9.8; refs. 25, 43). The present report complements TROPiCS-03, in which endometrioid histology was predominant and the patient population may have been less ethnically diverse. The current findings also align with the 22% ORR and OS of 11.9 months in patients with endometrial cancer who had relapsed after or were refractory to at least one prior standard treatment regimen enrolled in the phase I/II IMMU-132-01 basket trial (24). Currently, there are scant data available for patients with advanced endometrial cancer receiving Trop-2–targeting ADC treatments after progression on platinum-based therapy and ICI therapy. NCT04251416 thus provides the largest data to date in this setting, addressing a relevant unmet need.

In NCT04251416, we found that tumor Trop-2 protein was highly expressed in endometrioid and uterine serous carcinomas, with lower expression in carcinosarcoma patients (not shown). These IHC data are consistent with previously published preclinical work, in which endometrioid and uterine serous tumors were found to express Trop-2 in 98% and 95% of the specimens, respectively (22, 44), whereas only 30% of carcinosarcomas expressed Trop-2 (45). Carcinosarcomas may exhibit heterogeneity across carcinomatous and sarcomatous components (46), which could have implications for how this biomarker is used for this histology and an area for future inquiry. Importantly, in this study and previous clinical work (25), the benefit of SG was conferred even in tumors with lower Trop-2 H-scores, supporting the expanding roles ADCs can have in the treatment of patients with varied expression of their protein targets due to mechanisms such as bystander diffusion, high DAR, low receptor binding affinity, and extracellular cleavage of linkers (47) and as paralleled by agents such as trastuzumab deruxtecan with Her2 (48).

Only 50% of participants in this trial had been exposed to ICIs because evidence to include ICIs with carboplatin and paclitaxel emerged during its conduct; the original protocol was also conceptualized prior to the endorsement of pembrolizumab and lenvatinib as a preferred regimen for MMR-proficient recurrent disease by the National Comprehensive Cancer Network (NCCN; ref. 49). In the TROPICS-03 study, 85% of patients had prior exposure to ICIs (25). Trop-2 has been associated with primary resistance to ICIs (50), suggesting that anti–Trop-2 approaches may also occupy an important niche for patients who fail ICIs. An international phase III trial of SG versus physician’s choice chemotherapy (ASCENT-GYN-01/GOG-3104/ENGOT-en26/APGOT-EN2) in patients who have failed both platinum and ICI is currently underway (51). Irinotecan may also boost ICI-mediated T-cell activation through enhanced PDL1 expression (52), heralding a potential role for combinatorial strategies of SG and ICIs in endometrial cancers.

As the Trop-2 exploratory analyses were conducted in a relatively small cohort (n = 43), our observations warrant further corroboration in a larger study with a control arm more elegantly designed for the evaluation of Trop-2 and efficacy relationships and more extensively in the post-ICI or ICI-refractory setting. Tumors that overexpress Her2 have alternate therapeutic options by means of trastuzumab and trastuzumab deruxtecan; therefore, SG may represent a particularly important therapeutic strategy for Her2 non-overexpressing tumors. Additional study would be warranted to determine the optimal sequencing of SG and trastuzumab deruxtecan for patients with Her2 overexpression and recurrent disease. Sequential treatment with ADCs that target different antigens despite mechanistically similar payloads represents a viable approach but yields variable responses, underscoring the complex, multifactorial nature of ADC resistance and an emerging area for inquiry (53, 54).

Limitations of this study include the lack of a comparator arm, small sample size, and incompleteness of the exploratory data. These factors prevent meaningful correlation of AEs with Trop-2 expression, which has not been rigorously conducted in any phase III randomized study with SG. In the phase III ASCENT trial, safety in subgroups was consistent with the overall study population (41). Notably, the inclusion criteria evolved during stages 1 and 2 of the study. Importantly, the interim analysis between stages 1 and 2 showed no correlation between Trop-2 expression and response; the ORRs for stages 1 versus 2 were similar and both satisfied the primary endpoint (33% and 25%, respectively), even though the number of carcinosarcomas enrolled in stage 2 was more than double that of stage 1.

Conclusions

Findings from the phase II study NCT04251416 showed encouraging efficacy of SG, with a manageable toxicity profile in a heavily pretreated population which included uterine serous carcinoma, carcinosarcoma, and grade 3 endometrioid chemotherapy-resistant endometrial cancer. No new safety signals were uncovered.

Supplementary Material

Supplementary Data 1

Protocol.

Supplementary Data 2

Modified CONSORT diagram.

Supplementary Data 3

SAEs.

Supplementary Table 1

Representativeness of Study Participants.

Data Availability

Anonymized individual patient data/materials/protocols may be shared upon request or as required by law or regulation with qualified external researchers, if the request is made within 6 years of publication. Data requests should be sent to alessandro.santin@yale.edu.

Authors’ Disclosures

A.D. Santin reports grants and personal fees from Gilead during the conduct of the study, as well as grants from Merck, Daiichi Sankyo, and Verastem; personal fees from Eisai and Pfizer; and grants and personal fees from Gilead and R-Pharm US outside the submitted work. N. Buza reports personal fees from AstraZeneca, Daiichi Sankyo, AbbVie, Merck, GSK, and Diaceutics outside the submitted work. M. Clark reports personal fees from AbbVie and grants from the NIH outside the submitted work. No disclosures were reported by the other authors.

Authors’ Contributions

A.D. Santin: Conceptualization, data curation, supervision, writing–review and editing. D.M. Roque: Data curation, writing–original draft, writing–review and editing. E.R. Siegel: Data curation, formal analysis, validation, writing–original draft, writing–review and editing. V.M. Ettorre: Data curation, investigation, writing–review and editing. M. Greenman: Data curation, investigation, writing–review and editing. B. McNamara: Data curation, investigation, writing–review and editing. K. Papatla: Data curation, investigation, writing–review and editing. A. Kailasam: Data curation, investigation, writing–review and editing. N. Buza: Investigation, writing–review and editing. M. Clark: Data curation, validation, writing–review and editing. P. Hui: Investigation, writing–review and editing. P. Dottino: Data curation, investigation, writing–review and editing. E. Ratner: Data curation, investigation, writing–review and editing. S. Bellone: Conceptualization, resources, data curation, formal analysis, supervision, validation, investigation, visualization, methodology, writing–original draft, writing–review and editing.